NetXMS Support Forum

Victor could probably provide a patch to just fix this issue - where you wouldn't have to rebuild everything.

May be best to contact him directly.

I personally build and package the entire server almost daily - when there are new commits to the NetXMS GIT repo.
(not manually, I automated the build/package process)

We have an apt repo where I push the packages to.
We then update our system (and our customer systems) as required off that apt repo.

You can build from sources and it will work.
https://github.com/netxms/netxms/tree/stable-2.2

Previous events in the DB will stay without node set.
(not sure if there is a way to recover which nodes the events came from - would guess not)

As to when next release happens - I don't know, Victor will need to tell us

Just an update, this is now fixed in master and stable-2.2 branches.

The fix will also be out in the next release of NetXMS of course

Confirming this on using 2.2.3.

In syslog monitor, correct node is seen.
However, looking at event log on the node itself, the event is not assigned to the node.

Event does seem to enter EPP, but $node is not set, so using action with NXSL scripts that use $node generates an NXSL error.
(function or operation is not an object)

So it seems events are generated, just don't have NODE internally set to them.
They then correctly go to EPP, but since there is no NODE, alarms will not be generated on the node, NXSL scripts using $node will fail, etc.

For Trusted Nodes, by default, you should have "CheckTrustedNodes" set to "0" (false) in Server Configuration.
Trusted Nodes functionality is usually only needed in very secure environments.

More info here:
https://wiki.netxms.org/wiki/SG:Security_Issues

On the node used as the SSH proxy, do you have the SSH SubAgent declared properly?

Hi,

There are a few ways in which you can accomplish this:
1) ExternalParameter
2) ExternalParameterProvider
3) Use native SSH-based DCI
(assuming you take this as an output of some command from a SSH CLI of the device)

You can find a bunch of any on any of this on the Wiki, the Admin Guide, and here on the forums.

Feel free to ask if you have any questions.

"." is indeed string concatanation.

It's shown and explained (along with others base NXSL concepts) here:
https://wiki.netxms.org/wiki/UM:NetXMS_Scripting_Language_(NXSL)

"->" is used to access members of an object
(attributes or functions)

You can find more details here:
https://wiki.netxms.org/wiki/NXSL_Class_Reference

"error 15 in line 5: uknown object's attribute." means that the object you are referencing doesn't have the member you specified.

You should contact Raden (the company behind NetXMS) for a white-label license.

More info here:
https://www.radensolutions.com/

You should contact Raden (the company behind NetXMS) for a white-label license.

More info here:
https://www.radensolutions.com/

"localhost" might be getting resolved differently by nxdbmgr and MySQL.
(for example, one resolves an IPv4 and one an IPv6 address)

Try using "127.0.0.1" just to be sure.

Hi,

If you have the email address as one of the attributes of the users in LDAP, simple change the "LdapMappingName" parameter value in NetXMS server configuration to the name of that attribute.

When users are synced from LDAP into NetXMS, the attribute configured in "LdapMappingName" is used as the login name.

Quote from: jdowney on November 10, 2017, 01:11:36 PM
Should all agents in each zone have the Zone ID specified? Or just the machines that are being used as proxies?

Also am I correctly in thinking the proxy's cache Syslog and SNMP traps etc when a link to the master server is lost?

John

Only the proxy needs ZoneId configured.

I am actually not sure about caching - Viktor will be the best to answer that.

All proxy related settings are this:


# Proxying related settings
EnableProxy = yes
EnableSNMPProxy = yes
EnableSNMPTrapProxy = yes
EnableSyslogProxy = yes
ZoneId = 1017


You need to configure ZoneId for proper zone matching of Syslog and SNMP Traps.

Yes, syslog proxy has been fully implemented, zone matching is properly supported, etc.

We use Raspbian Lite on our Pis.

Your alarm key in the EPP rule that processes your event is wrong.

You need to make it per-node there.